A new way of thinking about cloud. Achievement of Zero Friction.
I recently participated in an interesting debate at the Gartner Catalyst event. ”The Battle for Cloud Hill: Public vs Private Cloud”
I think we’d all agree we are still in the early innings of a long game, and the debates and forums like this are helping to shape thought leadership and help businesses make decisions that will have direct impact in the short term, and also for the next 5, 15, or 50 years.
So, what does public and private mean? Are these terms clear cut, or even fair? Probably not. But from the debater’s chair position I had to take a side.
My debating partner @mayurshintre did a fantastic job representing the enterprise IT stakeholders point of view that leans toward private cloud - that this is a transitional period, and enterprise IT operated datacenter solutions aren’t going anywhere for a while.
That had me thinking more along the lines of ownership and operation - not the typical NISTy definitions of cloud that these debate lines have typically been cut with (e.g. public API, multi-tenancy, etc).
Instead of obsessing over things like multi-tenancy, and arguing over how cloud-washed virtualization isn’t really cloud, we should recognize that there is enough economic activity and a seismic shift happening and take a simpler approach.
I’d like to propose a new way of thinking about cloud that takes an empirical viewpoint and considers the economics around the entire the supply chain - from the finished good back to the earth beneath the datacenter - after all that is what everyone is worried about being disrupted.
Instead of black and white categorization, the cloud ecosystem, and a given organizations participation in that ecosystem, can be measured by it’s achievement of a zero-friction, on-time delivery supply chain for technology services, or simply, a zero-friction cloud.
There are two extremes in this continuum:
- Zero-Friction Clouds: AWS, SalesForce, Heroku
- Project Clouds: Virtualization environment running in a server closet, Private OpenStack deployment in an enterprise owned datacenter, running SugarCRM in a colo facility
There is nothing in this model that prevents an enterprise from having a zero-friction cloud of their own - or behaving like a service provider and implementing that type of supply chain - the largest companies that have competitive advantage to do so, will.
Zero Friction Clouds tend to be:
- service provider operated,
- consumed by one or many enterprises (or organizations, or consumers),
- with an interface bound supply chain - API’s, portals and their related systems
- driven by arms-length purchasing activity of raw resources (IaaS), intermediate services (PaaS), and finished goods (SaaS).
Project Clouds are:
- enterprise operated,
- consumed by that same enterprise,
- with a physical goods based supply chain - servers, storage, switches, licenses, or appliances, and all the labor related to that supply chain*
- driven by direct purchasing activity through traditional direct and wholesale supply chain - all raw materials fashioned into finished goods by the organization itself.
* They may expose API and self-service, but they are ultimately managed and operated on the same balance sheet as the businesses and people who consume them are on, and are bound by those resources
Now, you may say, “Project Clouds” are not really clouds at all, but we’ll concede that point to move beyond the debate. I believe this way of thinking can help IT stakeholders move toward achievement of the ideal vs. hanging on to the past.
I will be digging into this more. Stay tuned.